Answer Engine SEO: How to Make Your Website AI-Readable in 2025

People no longer search just to click - they ask to decide: “Where can I get cacio e pepe near the Colosseum right now? Do they have gluten-free?” Answer engines (ChatGPT, Gemini, Perplexity, Bing Copilot, Apple Intelligence) compile a single response from multiple sources using retrieval-augmented generation (RAG), vertical knowledge graphs, and structured snippets. If your site is easy to parse, you get included and cited. If not, you’re invisible. decide retrieval-augmented generation (RAG) This isn’t classical SEO theatre. It’s machine readability engineering. machine readability What Actually Happens Under the Hood (High-Level Pipeline) Real-world AI answer engines vary, but the architecture rhymes: Crawling & Fetching

Respect robots.txt, sitemaps, canonical URLs. Prefer static HTML or server-side rendered (SSR/SSG) content. Hydration-only JSON apps get less love unless prerendered.

Content Extraction

Boilerplate removal (nav, footers, cookie banners) using DOM heuristics (Readability-like algorithms), visual density, and CSS role/ARIA hints. Main content scoring via tag semantics ( , , , ), heading hierarchy, and text-to-link ratios.

Normalization

Language detection, de-duplication, canonicalization, currency/unit normalization, timezone resolution (critical for hours/menus/events).

Structuring

Parse schema.org JSON-LD/Microdata/RDFa. Promote entity slots (Restaurant, Menu, MenuSection, MenuItem, Offer, Price, HoursSpecification, GeoCoordinates). Build internal knowledge graph edges: (Dish) —servedAt→ (Restaurant) —locatedIn→ (Rome).

Indexing

Create sparse/dense embeddings per section/chunk (BM25 + vector embeddings). Store table-like data separately (menu items, prices) to answer structured queries quickly.

Retrieval at Query Time

Query understanding → retrieve top-k chunks via hybrid search (BM25 + cosine sim) + schema filters (e.g., @type=MenuItem). Assemble context windows with citations and attribution-friendly snippets.

Generation

LLM answers with citations; when structured slots exist, they’re privileged over plain text. If you have structured data, you win tie-breaks. Crawling & Fetching

Respect robots.txt, sitemaps, canonical URLs. Prefer static HTML or server-side rendered (SSR/SSG) content. Hydration-only JSON apps get less love unless prerendered. Crawling & Fetching Respect robots.txt, sitemaps, canonical URLs. Prefer static HTML or server-side rendered (SSR/SSG) content. Hydration-only JSON apps get less love unless prerendered. Respect robots.txt, sitemaps, canonical URLs. robots.txt Prefer static HTML or server-side rendered (SSR/SSG) content. Hydration-only JSON apps get less love unless prerendered. static HTML Content Extraction

Boilerplate removal (nav, footers, cookie banners) using DOM heuristics (Readability-like algorithms), visual density, and CSS role/ARIA hints. Main content scoring via tag semantics ( , , , ), heading hierarchy, and text-to-link ratios. Content Extraction Boilerplate removal (nav, footers, cookie banners) using DOM heuristics (Readability-like algorithms), visual density, and CSS role/ARIA hints. Main content scoring via tag semantics ( , , , ), heading hierarchy, and text-to-link ratios. Boilerplate removal (nav, footers, cookie banners) using DOM heuristics (Readability-like algorithms), visual density, and CSS role/ARIA hints. DOM heuristics Main content scoring via tag semantics ( , , , ), heading hierarchy, and text-to-link ratios. Normalization

Language detection, de-duplication, canonicalization, currency/unit normalization, timezone resolution (critical for hours/menus/events). Normalization Language detection, de-duplication, canonicalization, currency/unit normalization, timezone resolution (critical for hours/menus/events). Language detection, de-duplication, canonicalization, currency/unit normalization, timezone resolution (critical for hours/menus/events). Structuring

Parse schema.org JSON-LD/Microdata/RDFa. Promote entity slots (Restaurant, Menu, MenuSection, MenuItem, Offer, Price, HoursSpecification, GeoCoordinates). Build internal knowledge graph edges: (Dish) —servedAt→ (Restaurant) —locatedIn→ (Rome). Structuring Parse schema.org JSON-LD/Microdata/RDFa. Promote entity slots (Restaurant, Menu, MenuSection, MenuItem, Offer, Price, HoursSpecification, GeoCoordinates). Build internal knowledge graph edges: (Dish) —servedAt→ (Restaurant) —locatedIn→ (Rome). Parse schema.org JSON-LD/Microdata/RDFa. schema.org JSON-LD schema.org Promote entity slots (Restaurant, Menu, MenuSection, MenuItem, Offer, Price, HoursSpecification, GeoCoordinates). entity slots Build internal knowledge graph edges: (Dish) —servedAt→ (Restaurant) —locatedIn→ (Rome). knowledge graph edges Indexing

Create sparse/dense embeddings per section/chunk (BM25 + vector embeddings). Store table-like data separately (menu items, prices) to answer structured queries quickly. Indexing Create sparse/dense embeddings per section/chunk (BM25 + vector embeddings). Store table-like data separately (menu items, prices) to answer structured queries quickly. Create sparse/dense embeddings per section/chunk (BM25 + vector embeddings). sparse/dense embeddings Store table-like data separately (menu items, prices) to answer structured queries quickly. Retrieval at Query Time

Query understanding → retrieve top-k chunks via hybrid search (BM25 + cosine sim) + schema filters (e.g., @type=MenuItem). Assemble context windows with citations and attribution-friendly snippets. Retrieval at Query Time Query understanding → retrieve top-k chunks via hybrid search (BM25 + cosine sim) + schema filters (e.g., @type=MenuItem). Assemble context windows with citations and attribution-friendly snippets. Query understanding → retrieve top-k chunks via hybrid search (BM25 + cosine sim) + schema filters (e.g., @type=MenuItem). top-k chunks schema filters @type=MenuItem Assemble context windows with citations and attribution-friendly snippets. citations attribution-friendly snippets Generation

LLM answers with citations; when structured slots exist, they’re privileged over plain text. If you have structured data, you win tie-breaks. Generation LLM answers with citations; when structured slots exist, they’re privileged over plain text. If you have structured data, you win tie-breaks. LLM answers with citations; when structured slots exist, they’re privileged over plain text. If you have structured data, you win tie-breaks. privileged over plain text The theme: structure beats prose. structure beats prose The Algorithms & Signals That Matter (and How to Feed Them) 1) DOM Semantics → Main Content Scoring Many extractors score blocks by: Heading depth ( near the top, reasonable / nesting). Semantic containers ( , , ). Density (characters per block, lower link density, fewer repeated patterns). ARIA roles (role="main", role="article"). Heading depth ( near the top, reasonable / nesting). Semantic containers ( , , ). Density (characters per block, lower link density, fewer repeated patterns). ARIA roles (role="main", role="article"). role="main" role="article" Do: Do: Trattoria Aurelia — Roman Classics Since 1978 Handmade pasta, seasonal produce, and late-night hours near the Colosseum.

Menu Cacio e Pepe Fresh tonnarelli with Pecorino Romano DOP and Tellicherry pepper.

€12

Trattoria Aurelia — Roman Classics Since 1978 Handmade pasta, seasonal produce, and late-night hours near the Colosseum.

Menu Cacio e Pepe Fresh tonnarelli with Pecorino Romano DOP and Tellicherry pepper.

€12

Don’t: nest the entire page inside with custom class names only. Models lose strong hints about what’s important. Don’t: 2) Structured Data (schema.org) → Entity Extraction schema.org LLM pipelines elevate JSON-LD above free text. Use Restaurant + Menu + MenuItem + Offer + OpeningHoursSpecification. Include sameAs, geo, priceRange, servesCuisine, acceptsReservations. Restaurant Menu MenuItem Offer OpeningHoursSpecification sameAs geo priceRange servesCuisine acceptsReservations Full Menu Example (JSON-LD) Full Menu Example (JSON-LD) { "@context": "https://schema.org", "@type": "Restaurant", "name": "Trattoria Aurelia", "servesCuisine": ["Italian", "Roman"], "priceRange": "€€", "telephone": "+39 06 5555 1234", "address": { "@type": "PostalAddress", "streetAddress": "Via dei Fori Imperiali 12", "addressLocality": "Rome", "addressRegion": "RM", "postalCode": "00184", "addressCountry": "IT" }, "geo": {"@type": "GeoCoordinates", "latitude": 41.8902, "longitude": 12.4922}, "acceptsReservations": "Yes", "openingHoursSpecification": [ {"@type": "OpeningHoursSpecification", "dayOfWeek": ["Monday","Tuesday","Wednesday","Thursday","Friday"], "opens": "11:30", "closes": "23:00", "validFrom": "2025-01-01"}, {"@type": "OpeningHoursSpecification", "dayOfWeek": ["Saturday","Sunday"], "opens": "11:30", "closes": "01:00"} ], "menu": { "@type": "Menu", "name": "Dinner Menu — Summer", "hasMenuSection": [ { "@type": "MenuSection", "name": "Pasta", "hasMenuItem": [ { "@type": "MenuItem", "name": "Cacio e Pepe", "description": "Tonnarelli, Pecorino Romano DOP, black pepper", "offers": {"@type": "Offer", "priceCurrency": "EUR", "price": 12} }, { "@type": "MenuItem", "name": "Amatriciana", "description": "Guanciale, tomato, Pecorino Romano", "offers": {"@type": "Offer", "priceCurrency": "EUR", "price": 13} } ] } ] }, "sameAs": [ "https://maps.google.com/?cid=...", "https://www.instagram.com/trattoriaaurelia", "https://www.facebook.com/trattoriaaurelia" ] } { "@context": "https://schema.org", "@type": "Restaurant", "name": "Trattoria Aurelia", "servesCuisine": ["Italian", "Roman"], "priceRange": "€€", "telephone": "+39 06 5555 1234", "address": { "@type": "PostalAddress", "streetAddress": "Via dei Fori Imperiali 12", "addressLocality": "Rome", "addressRegion": "RM", "postalCode": "00184", "addressCountry": "IT" }, "geo": {"@type": "GeoCoordinates", "latitude": 41.8902, "longitude": 12.4922}, "acceptsReservations": "Yes", "openingHoursSpecification": [ {"@type": "OpeningHoursSpecification", "dayOfWeek": ["Monday","Tuesday","Wednesday","Thursday","Friday"], "opens": "11:30", "closes": "23:00", "validFrom": "2025-01-01"}, {"@type": "OpeningHoursSpecification", "dayOfWeek": ["Saturday","Sunday"], "opens": "11:30", "closes": "01:00"} ], "menu": { "@type": "Menu", "name": "Dinner Menu — Summer", "hasMenuSection": [ { "@type": "MenuSection", "name": "Pasta", "hasMenuItem": [ { "@type": "MenuItem", "name": "Cacio e Pepe", "description": "Tonnarelli, Pecorino Romano DOP, black pepper", "offers": {"@type": "Offer", "priceCurrency": "EUR", "price": 12} }, { "@type": "MenuItem", "name": "Amatriciana", "description": "Guanciale, tomato, Pecorino Romano", "offers": {"@type": "Offer", "priceCurrency": "EUR", "price": 13} } ] } ] }, "sameAs": [ "https://maps.google.com/?cid=...", "https://www.instagram.com/trattoriaaurelia", "https://www.facebook.com/trattoriaaurelia" ] } Why it works: Retrieval systems can slot-answer queries like “price of cacio e pepe at Trattoria Aurelia” without re-parsing prose. Why it works: “price of cacio e pepe at Trattoria Aurelia” 3) Content Chunking → Embedding Recall LLMs use chunked embeddings. Oversized pages hurt recall; too many tiny chunks lose context. chunked embeddings Practice: Practice: Aim for 300–800 tokens per logical section (roughly 1–4 paragraphs + a table). Use headings and id anchors; keep each dish/FAQ in its own subsection. Provide permalink anchors (e.g., /menu#cacio-e-pepe) so retrievers can cite precise spans. Aim for 300–800 tokens per logical section (roughly 1–4 paragraphs + a table). 300–800 tokens per logical section Use headings and id anchors; keep each dish/FAQ in its own subsection. id Provide permalink anchors (e.g., /menu#cacio-e-pepe) so retrievers can cite precise spans. permalink anchors /menu#cacio-e-pepe Cacio e Pepe ...

Cacio e Pepe ...

4) Tabular Data → Parseable Tables, Not Pictures If prices/hours appear in tables, keep them as semantic tables or lists. Avoid images/PDFs. semantic tables Dinner Prices Dish Price (EUR) Allergens Cacio e Pepe 12 Milk, Gluten Amatriciana 13 Milk, Pork, Gluten Dinner Prices Dish Price (EUR) Allergens Cacio e Pepe 12 Milk, Gluten Amatriciana 13 Milk, Pork, Gluten Bonus: mirror the data in JSON-LD so both text and structure exist. Bonus: 5) Multilingual & Locale Signals → Correct Matching Query: “Where to eat cacio e pepe near the Colosseum, up to 15 euros?” Make sure your page clarifies language, currency, and timezone. language, currency, and timezone Use lang on and hreflang for alternates. Normalize currency with ISO codes in JSON-LD (priceCurrency: "EUR"). Provide local phone formats and openingHoursSpecification with validFrom/Through if seasonal. Use lang on and hreflang for alternates. lang hreflang Normalize currency with ISO codes in JSON-LD (priceCurrency: "EUR"). priceCurrency: "EUR" Provide local phone formats and openingHoursSpecification with validFrom/Through if seasonal. local phone formats openingHoursSpecification validFrom/Through 6) Canonicalization, Sitemaps, and Change Hints Retrievers prioritize freshness for operational facts (hours, prices, availability). Add lastmod to sitemaps and keep it honest. Use to avoid duplicate menus across UTM’d pages. Publish invalidation-friendly URLs: /menu/summer-2025 rather than /menu?date=1699. Avoid JS-only rendering for price text; SSR it. Add lastmod to sitemaps and keep it honest. lastmod sitemaps Use to avoid duplicate menus across UTM’d pages. Publish invalidation-friendly URLs: /menu/summer-2025 rather than /menu?date=1699. invalidation-friendly URLs /menu/summer-2025 /menu?date=1699 Avoid JS-only rendering for price text; SSR it. Non‑Trivial, Production‑Level Patterns A) Disambiguate Similar Entities with @id and Stable Anchors @id If you have two venues (Trastevere vs. Monti), give each a stable @id and specific geo. @id geo { "@context": "https://schema.org", "@id": "https://trattoriaaurelia.it/locations/monti#restaurant", "@type": "Restaurant", "name": "Trattoria Aurelia — Monti", "geo": {"@type": "GeoCoordinates", "latitude": 41.894, "longitude": 12.494}, "address": {"@type": "PostalAddress", "addressLocality": "Rome"} } { "@context": "https://schema.org", "@id": "https://trattoriaaurelia.it/locations/monti#restaurant", "@type": "Restaurant", "name": "Trattoria Aurelia — Monti", "geo": {"@type": "GeoCoordinates", "latitude": 41.894, "longitude": 12.494}, "address": {"@type": "PostalAddress", "addressLocality": "Rome"} } Now an AI can answer: “Which branch serves cacio e pepe past midnight?” by joining MenuItem with the Monti branch’s hours. “Which branch serves cacio e pepe past midnight?” MenuItem Monti B) Recipe/Allergen Knowledge → DietaryRestriction & SuitableForDiet DietaryRestriction SuitableForDiet { "@context": "https://schema.org", "@type": "MenuItem", "name": "Cacio e Pepe", "suitableForDiet": ["https://schema.org/VegetarianDiet"], "menuAddOn": [ {"@type": "MenuItem", "name": "Gluten-free pasta", "offers": {"@type": "Offer", "price": 2, "priceCurrency": "EUR"}} ], "requiresSubscription": false } { "@context": "https://schema.org", "@type": "MenuItem", "name": "Cacio e Pepe", "suitableForDiet": ["https://schema.org/VegetarianDiet"], "menuAddOn": [ {"@type": "MenuItem", "name": "Gluten-free pasta", "offers": {"@type": "Offer", "price": 2, "priceCurrency": "EUR"}} ], "requiresSubscription": false } This lets assistants answer “Is their cacio e pepe vegetarian? Can I get gluten-free?” without hallucinating. “Is their cacio e pepe vegetarian? Can I get gluten-free?” C) Temporal Facts → Validity Windows Menus change. Model pipelines weight recent structured facts. recent { "@type": "Offer", "price": 12, "priceCurrency": "EUR", "priceValidUntil": "2025-10-01" } { "@type": "Offer", "price": 12, "priceCurrency": "EUR", "priceValidUntil": "2025-10-01" } Pair with sitemap.xml updates so freshness signals align. sitemap.xml D) FAQ Blocks → Extractive Answer Boosters LLMs love Q&A pairs. Provide FAQPage with concise answers and link to canonical detail. FAQPage { "@context": "https://schema.org", "@type": "FAQPage", "mainEntity": [ { "@type": "Question", "name": "Do you accept walk-ins after 11pm?", "acceptedAnswer": {"@type": "Answer", "text": "Yes, until 12:30am on weekends."} } ] } { "@context": "https://schema.org", "@type": "FAQPage", "mainEntity": [ { "@type": "Question", "name": "Do you accept walk-ins after 11pm?", "acceptedAnswer": {"@type": "Answer", "text": "Yes, until 12:30am on weekends."} } ] } E) Robust Images → ALT, Figure Captions, and EXIF Hygiene Use … . Keep filenames descriptive (cacio-e-pepe-pecorino.jpg). Don’t bake text into images; assistants can’t extract it reliably. Use … . … Keep filenames descriptive (cacio-e-pepe-pecorino.jpg). cacio-e-pepe-pecorino.jpg Don’t bake text into images; assistants can’t extract it reliably. F) Events & Reservations → Event + Deep Links Event { "@context": "https://schema.org", "@type": "Event", "name": "Truffle Week", "startDate": "2025-11-02", "endDate": "2025-11-09", "location": {"@type": "Place", "name": "Trattoria Aurelia — Monti"}, "offers": {"@type": "Offer", "url": "https://trattoriaaurelia.it/reserve?event=truffle-week"} } { "@context": "https://schema.org", "@type": "Event", "name": "Truffle Week", "startDate": "2025-11-02", "endDate": "2025-11-09", "location": {"@type": "Place", "name": "Trattoria Aurelia — Monti"}, "offers": {"@type": "Offer", "url": "https://trattoriaaurelia.it/reserve?event=truffle-week"} } Assistants can now answer “Any truffle events in Rome next week?” and deep-link bookings. “Any truffle events in Rome next week?” How This Maps to RAG & Embeddings Hybrid Retrieval: Systems combine classical IR (BM25) with dense vectors. Your job: make exact tokens discoverable (dish names, neighborhoods, hours) and provide rich context for vectors. Chunk Boundaries: Use headings and logical grouping; don’t interleave unrelated content in one paragraph (e.g., lunch and dinner prices together). Anchor-able Citations: Provide stable anchors so assistants can attribute precisely - this improves the chance you’re cited. Hybrid Retrieval: Systems combine classical IR (BM25) with dense vectors. Your job: make exact tokens discoverable (dish names, neighborhoods, hours) and provide rich context for vectors. Hybrid Retrieval: and Chunk Boundaries: Use headings and logical grouping; don’t interleave unrelated content in one paragraph (e.g., lunch and dinner prices together). Chunk Boundaries: Anchor-able Citations: Provide stable anchors so assistants can attribute precisely - this improves the chance you’re cited. Anchor-able Citations: Internal Linking = Graph Strength Think in graphs: link dish → origin story → ingredient sourcing → allergen policy. Each link with a descriptive anchor boosts entity resolution and gives retrievers semantic hops. Pitfalls That Break AI Parsing (Seen in Production) Menus as PDFs/JPEGs - invisible. Provide HTML + JSON-LD mirror. Prices rendered only client-side - crawlers time out on hydration or block XHR; SSR your critical content. Infinite scroll for core info - put the essentials above-the-fold in the initial HTML. Locale confusion - 12,00 vs 12.00, ambiguous timezones. Normalize in JSON-LD. Over-nested div soup - no semantic hints; extractors miss the main content. Duplicated pages without canonical - embeddings split authority; citations fragment. Menus as PDFs/JPEGs - invisible. Provide HTML + JSON-LD mirror. Menus as PDFs/JPEGs Prices rendered only client-side - crawlers time out on hydration or block XHR; SSR your critical content. Prices rendered only client-side Infinite scroll for core info - put the essentials above-the-fold in the initial HTML. Infinite scroll for core info Locale confusion - 12,00 vs 12.00, ambiguous timezones. Normalize in JSON-LD. Locale confusion 12,00 12.00 Over-nested div soup - no semantic hints; extractors miss the main content. Over-nested div soup Duplicated pages without canonical - embeddings split authority; citations fragment. Duplicated pages without canonical A Concrete Checklist set; localized alternates with hreflang. SSR/SSG for critical text (menu, hours, address, phone). , , correct H1/H2 hierarchy. schema.org JSON-LD for your vertical (Restaurant, Product, Event, FAQ, Review). Each item (dish, product) gets its own and anchor link. Tables are , not screenshots. Sitemap with lastmod; stable canonical URLs. Use @id, sameAs, geo, openingHoursSpecification. Provide allergens and dietary flags. Test with multiple parsers (Google Rich Results test + a Readability clone + your own micro RAG script). set; localized alternates with hreflang. hreflang SSR/SSG for critical text (menu, hours, address, phone). , , correct H1/H2 hierarchy. schema.org JSON-LD for your vertical (Restaurant, Product, Event, FAQ, Review). schema.org Each item (dish, product) gets its own and anchor link. Tables are , not screenshots. Sitemap with lastmod; stable canonical URLs. lastmod Use @id, sameAs, geo, openingHoursSpecification. @id sameAs geo openingHoursSpecification Provide allergens and dietary flags. Test with multiple parsers (Google Rich Results test + a Readability clone + your own micro RAG script). Bonus: Build a Tiny RAG Parser to “Think Like an Answer Engine” You can locally simulate how assistants will see your page: Fetch your rendered HTML (SSR). Run Readability (or a Node clone) to extract main content. Parse JSON-LD via a micro schema.org extractor. Chunk by headings; embed each chunk (any open-source SentenceTransformer). Ask queries; retrieve top-k; verify the text is sufficient without images/JS. Fetch your rendered HTML (SSR). Run Readability (or a Node clone) to extract main content. Readability Parse JSON-LD via a micro schema.org extractor. schema.org Chunk by headings; embed each chunk (any open-source SentenceTransformer). Ask queries; retrieve top-k; verify the text is sufficient without images/JS. Once your page answers your own local RAG, you’ve cleared the biggest hurdle. Other Vertical Playbooks E-commerce: Product, Offer, AggregateRating, Review, ItemAvailability, Brand, GTIN, color/size variants as separate @id nodes. Events: Event with startDate/endDate, eventAttendanceMode, location.geo, ticket Offer with currency and availabilityEnds. Local Services: LocalBusiness with Service items (areaServed, hasOfferCatalog). Docs & APIs: TechArticle, HowTo, stable permalinks per endpoint, code blocks labeled, and FAQPage for common errors. E-commerce: Product, Offer, AggregateRating, Review, ItemAvailability, Brand, GTIN, color/size variants as separate @id nodes. E-commerce Product Offer AggregateRating Review ItemAvailability Brand GTIN color/size @id Events: Event with startDate/endDate, eventAttendanceMode, location.geo, ticket Offer with currency and availabilityEnds. Events Event startDate/endDate eventAttendanceMode location.geo Offer availabilityEnds Local Services: LocalBusiness with Service items (areaServed, hasOfferCatalog). Local Services LocalBusiness Service areaServed hasOfferCatalog Docs & APIs: TechArticle, HowTo, stable permalinks per endpoint, code blocks labeled, and FAQPage for common errors. Docs & APIs TechArticle HowTo stable permalinks per endpoint FAQPage Metrics and Monitoring Track assistant referrals (UTMs per assistant, e.g., ?ref=assistant-chatgpt). Measure crawl frequency via log analysis (user-agents like “GPTBot”, “CCBot”, “PerplexityBot”, etc.). Watch index freshness: time from deploy → assistant mentions updated price. A/B test schema richness on a subsection of pages. Track assistant referrals (UTMs per assistant, e.g., ?ref=assistant-chatgpt). assistant referrals ?ref=assistant-chatgpt Measure crawl frequency via log analysis (user-agents like “GPTBot”, “CCBot”, “PerplexityBot”, etc.). crawl frequency Watch index freshness: time from deploy → assistant mentions updated price. index freshness A/B test schema richness on a subsection of pages. schema richness If you can’t measure whether assistants picked you as a source, you’re guessing. If you can’t measure whether assistants picked you as a source, you’re guessing. Ethical & Practical Notes Respect accessibility: what helps screen readers helps machines. Avoid deceptive markup - penalties are coming. Publish privacy-respecting structured data (no dark patterns in JSON-LD). Respect accessibility: what helps screen readers helps machines. Avoid deceptive markup - penalties are coming. Publish privacy-respecting structured data (no dark patterns in JSON-LD). privacy-respecting Semantic Markup as an AI Contract Semantic HTML and structured data are not cosmetics; they’re a contract with machine readers. When you honor it - clear semantics, honest structure, fresh metadata - you’re easy to retrieve, easy to cite, and hard to hallucinate. contract Good markup today is discoverability tomorrow. Ship it.